-
Create a new GitHub repository from this template: Click on this link to create a new repository that has the same folder structure as our template. Make sure that you create a public repository.
- Alternatively, you can also use the other public alternatives that binder supports. For example, your own hosted public Git repository or a public GitLab repository.
-
Use on any data you want:
- We recommend accessing data files using web APIs (e.g. with the
requests
package). This especially holds for larger datasets. If you access the data files frequently, you might want to host them yourself. - If you have small datasets (< 1 MB in total), you can copy them into the
data_for_container/
folder (you can use the.parquet
file format to compress your files. pandas has methods for reading and writing parquet files). - You can also create your own docker image including all your data and use that as a base in the
Dockerfile
.
- We recommend accessing data files using web APIs (e.g. with the
-
Add own Notebooks: copy your notebook files in the
data_for_container/
. -
Create your own binder link / button: Go to https://mybinder.org/ and fill in your repository's details.
-
Share the binder link with your friends and the world.